Droping English_QuestionText as it doesn't have any information

Dropping Textbox19 feature as it doesn't have any valuable information

Renaming Features names

Checking null values

Droping NaN rows

Extracting Year and Month information from the data column

There is no relation between reference number and feedback. So removing it for now

Creating new features according to the information from the services features like Organisation, GP, priority area and so on

Data Cleansing and Data Wrangling

Apply a first round of text cleaning techniques

Expanding contraction words like didn't to did not

Apply a second round of text cleaning techniques

Removing Whitespaces

Finding Duplicates

Exploratory Data Analysis

Count for Organisation NHS Bradford CCG

Count for Organisation Yorkshire Ambulance Service

Count for Organisation Eccleshill Treatment Care

Count for Organisation Airedale Hospital

Count for Organisation Bradford District Care

Count for Organisation Bradford Teaching Hospital

Count for Organisation Optegra

Optegra has only one feedback so not proceeding with the data visualisation

Count for Organisation The Yorkshire Clinic

Count for Organisation The Yorkshire Eye Clinic

Text pre-processing

Tokenization, POS tagging, Stopwords removal

Choosing stopword technique

NLTK - STOPWORDS

SPACY - STOPWORDS

GENSIM - STOPWORDS

SKLEARN - STOPWORDS

After compared different stopwords techniques from nltk,spacy,gensim, and sklearn . I could see the more stopwords removed when I use SPACY . So I prefer to use spacy to remove stopwords from the text

Cleaned feedback with out the stop words

Lemmatization

After adding the custom stop words

Second round of stop word removal

creating wordcloud to check the words

POS tagged dictionary

WordCloud

Bi-Gram build and Word Cloud

Sentiment Analysis

Sentiment using spacy

Sentiment Visualisation

Word Cloud for sentiment

Word Cloud for spacy sentiment

Bi-gram Word Cloud for spacy sentiment

Finding most Common Words

Spacy 1-gram most common words

Positive Words

Negative Words

Spacy Bi-gram most common words

Positive Words

Negative Words

Sentiment Analysis based on the Organisation

Organisation NHS Bradford CCG

Organisation Airedale

Organisation District Care Hospital

Organisation Teaching Hospital

Organisation Ambulance Service

Organisation Yorkshire Clinic Service

Organisation ETC

Organisation Eye Clinic

Organisation Optegra

Latent Dirichlet Allocation - Topic Modelling

Spacy Combo

Dominant topic and its percentage contribution in each document

most representative sentence for each topic

#Decided Topics for this model,

Topic 1.0 - Appointment Issues

Topic 4.0 - General Complaints

Topic 5.0 - Mental health service issues

Topic 11.0 - Access Issues

Topic 14.0 - Carer/Interpretter issues

Exploratory Analysis - Sentiment Analysis

Ambulance Service

Teaching Hospital

District Care

Airedale Hospital

Topic Allocation

Added Negative comment Analysis

CCG

Teaching Hospital

District Care

Specific to services

Teaching Hospital

District Care